Button problem causing software malfunction


long time OT user, but new to the forum - I hope that I post in the correct section.

I have a problem, and I wonder if someone has encountered anything similar and/or perhaps could help me with some insight how to fix the problem or move forward.

My problem
1 The button “Exit/no” only reads a pressing on it 20-50 % of the times I press it. This phenomena started after I sensed a shift in the button, during normal use at home - now, the button also has a clicky element to it and it feels misaligned.
---------- I understand that this is not uncommon for Octatracks, and I could probably fix this with the help of existing sources on the web.

2 The OT works fine in general, and every other button checks out fine.

3 This is the annoying part: after having made the exit/no-button register being pressed, the channel select buttons (T1-8) stops working. The work fine prior to pressing the exit/no-button.
---------- I wonder if the exit/no-button sends a continuous message after being pressed?
---------- Is it possible that the faulty button somehow also causes a software problem?

If anyone could help me I would be very grateful. Regardless of this, my goal is to post the solution I reach, here in this tread.


Possible explanation for track select not working (unlikely but worth sharing?) :

Could you be holding down the track select buttons too long? I think this is intended behaviour to allow you to start a track recorder, or stop a sample without shifting focus to that track.

One time I was concerned about the track buttons & I kept holding down even longer because I wanted a ‘definitive’ press for testing! As soon as I returned to the more natural quick presses, they worked.


A sticky button could be one that is physically catching (less likely) has gunk/residue in it (but is otherwise unbroken) or is internally damaged

Fixes for the former two are searchable on here and are within a normal users’ reach - fixes for the latter involve a confident solderer and a spare part

I’m not sure if it helps with a sticky button, but it may be worth trying the test mode to see what it reports

Alternatively search the list of key combos for one using No+[…] to see if it is clearly sticking, I’m sure there’ll be one which nails whether it’s just not returning quick enough

If you search on removing the button cap safely on these pages you may be able to ascertain if the switch is gunky or broken


I’m willing to bet the case of the exit/no switch has cracked open (I’ve had several do this). If you’re really careful you can open up the machine and attempt to reseal the switch case with super glue, but this can be pretty tricky if you’re not aware of the switch mechanics. Otherwise you might need to send it in for service


Hi everyone, thanks for the input.

The problem is fixed now, and I will describe how I solved it, but first:

[@]Clancy thanks for the tip, and I wish that this had been the case all along!

[@]avantronica I would say that the button was mechanically misaligned. As for the test mode, since I did not know the correct test response for the buttons it was hard to draw any decisive conclusions from it. Having removed the button cap, it was easier to ascertain that the button mechanism was faulty.

[@]panelist you are spot on, sire.

step by step solution
1- I removed the button cap to inspect the button mechanism, and I found that it seemed “misaligned”. Instead of the spring mechanism under the button being a perfectly vertical pillar, it had an angle to it.

2- I opened up the machine to get a better look (hey the warranty’s out anyway…)

Here are some good links that boosted my confidence for doing this:
good video showing how to open an octatrack
[origin of the video:
https://www.reddit.com/r/synthesizers/comments/6fc0bj/repairing_my_octatrack_encoder/ ]

3- the “housing” for the button mechanism had come loose. Its “foundation” was secured to the pcb, but the housing on top of it had come loose on two corners, thus hanging somewhat precariously to its connection with the board.

4- much like panelist said, it was possible to correct this with super glue.

I constructed glue application tools from a toothpick that I made shavings of, and I bent a little hook in one end of them, making them look like tiny hockey sticks.

The glue I used was hobbyist glue for plastic and metal - i.e. clear and runny glue, not amber coloured epoxy glue (which is more brittle, not so good with plastic, and at best super messy).

I carefully applied the glue with my little hockey stick, whilst pressing down the housing on its foundation. Important: I continually exercised the button as the glue set, in order to avoid accidentaly gluing the button and/or any of the spring mechanism into place.

[I was going to make some links to other forum treads where panelist chimes in, but I’m not allowed as I am a new user…]

5- after re-assembly, the machine works fine again. My guess is that the faulty exit/no-button caused some kind of interference with the track select buttons. Thankfully (for now!) it works just like it should.

6- what’s my skill level? A few years of playing (old) Warhammer has made me pretty nifty with super glue. I have a modular system, making me feel confident with ribbon cables and circuit boards (but hey, it’s not that complicated). I have no soldering skills whatsoever.


These buttons are a terrible design - they fail closed and make the instrument unusable.

I’ve just had to pull my ‘cue’ button as I need the machine for a show in a week.

Upon opening the machine I noticed the Time Change, second Rec, and Up buttons are all loose and due to fail soon - fixed these (hopefully!) using super glue. Also the Play button, which I wasn’t game enough to tamper with.

There is a place here that replaces the buttons, ~$90 each, which isn’t unreasonable, but in time that’ll be $90 times… a fair few buttons.

1 Like

I think that is actually completely unreasonable.
With the proper tools it takes a couple of minutes per button. The buttons are available as a spare part from our support and any local professional with a de-soldering gun should be able to do it. A minimum price for the job is understandable, but 90$ per button sounds strange.


I’ve got mine taken apart right now. One of the buttons was raised from its base. I will superglue it back down, but I want to be sure of the button mechanics? Does the spring go to the back of the shaft with the bearing resting directly on the contacts? I don’t understand how the spring has any effect on the switch if the bearing is always resting on it. Anyone who has done this repair and can chime in would be appreciated.

Bench fees are usually $50-$100 for a competent professional, so it’s more like a reasonable bench fee plus about $20 for parts and labor.

But yeah, it’s really expensive, you could get a decent soldering station and supplies (solder wick, solder, etc.) for the cost of replacing one button and then be all set to replace them at cost any time they failed in the future. Buttons aren’t the easiest component to replace because of the hassle of desoldering and loosening all four legs but they’re not bad. The Octatrack might not be the best thing to learn on but I’d say with a little bit of practice (find an old broken radio or something and desolder all of the switches and jacks and you should have enough experience to swap a switch or two in the OT) it’s no big deal and it will save a lot of hassle and money to learn it.

Having done both over the years, replacing a faulty switch is almost always easier than trying to repair it, and if you’ve got the dexterity to do a repair like the one described above then soldering should be no problem at all.